Case Studies / Fr. Meyer’s Sohn
LOGISTICS · GLOBAL OPERATIONS
Automating information extraction from multilingual logistics emails
Global logistics company Fr. Meyer’s Sohn needed to extract structured operational data from thousands of unstructured emails in German and English. Traditional approaches couldn’t handle the variety. We built a GPT-powered extraction pipeline deployed on their own servers.
Client: Fr. Meyer’s Sohn – international logistics and shipping company with global operations.
KEY RESULTS
2 languages
German & English processed automatically
-80%
Reduction in manual extraction effort
PoC → Prod
From proof of concept to production deployment
On-premise
Deployed on client’s own servers via Docker
INDUSTRY
Logistics & Shipping
USE CASE
Email data extraction
AI APPROACH
GPT-3.5 / GPT-4 + prompt engineering
DEPLOYMENT
Docker on client servers
LANGUAGES
German & English
ENGAGEMENT
PoC → Production build
The challenge
Fr. Meyer’s Sohn is a global logistics company processing large volumes of operational emails daily. These emails contain critical logistics data – shipment details, routing information, scheduling data – but arrive in completely unstructured formats, written in both German and English.
The company needed to extract specific, predefined information from these emails and output it in a structured format for downstream processing. The challenge was significant: no two emails looked the same. Formats varied by sender, country, and language. Not all required fields were always present. Local standards for dates, addresses, and reference numbers differed across markets.
Traditional rule-based extraction approaches couldn’t handle the variety. The unstructured, multilingual nature of the content meant that any approach relying on fixed templates or keyword matching would fail at the scale Fr. Meyer’s Sohn operates.
What we built
We started with a proof of concept to demonstrate that generative AI could handle the extraction task reliably. After the client validated the results, we moved to a full production build.
GPT-powered extraction pipeline. We built an extraction system using GPT-3.5 and GPT-4, with carefully engineered prompts optimized for logistics-specific data. The system reads incoming email text, identifies the relevant data points specified by the client, and returns them in a clean, structured format.
Multilingual handling. The system processes German and English content natively, handling the differences in formatting, date conventions, and terminology that come with global logistics operations. Missing fields are flagged rather than guessed, maintaining data integrity.
Production-grade deployment. The solution was built with FastAPI and packaged in a Docker container for deployment on the client’s own servers. Fr. Meyer’s Sohn sends requests to the API and receives structured data back – fully integrated into their existing operational workflow.
Error handling and monitoring. Production-ready logging, error handling, and maintenance tooling were built in from the start – ensuring the system runs reliably at scale without requiring constant attention.
The results
BEFORE
Manual extraction from unstructured emails. Time-consuming, error-prone, and impossible to scale with growing email volume.
AFTER
Automated extraction via API. Structured data delivered in real time, multilingual support, deployed on client’s own infrastructure.
The automated system now efficiently identifies logistics-relevant information from unstructured emails, regardless of language or format. Manual extraction effort was reduced dramatically, and the structured output integrates directly into Fr. Meyer’s Sohn’s operational processes.
The flexible architecture allows the client to quickly adapt the system to new data requirements or additional languages as their operations evolve.
Technology used
More Case Studies
See how we’ve helped other companies
AUTOMOTIVE (UNDER NDA)
Replacing reactive maintenance with a predictive maintenance roadmap
Unplanned equipment failures were causing costly production stops. We assessed the manufacturer’s infrastructure and delivered a concrete architecture for predictive maintenance, a roadmap from reactive repairs to data-driven prevention.
PdM
Architecture
Roadmap
Delivered
MANUFACTURING · RADAWAY
Making email-based order processing reliable with LLMs
Staff were manually reading customer emails, identifying products, and entering orders by hand. We turned a promising AI prototype into a production system that handles it end to end, across languages, formats, and attachments.
-90%
Manual intervention
95%+
Match accuracy
MEDTECH · APOQLAR
Cutting security questionnaire completion from one month to one week
Every new hospital client required a completed questionnaire, 8–10 people, about a month, pulling answers from policies across departments. Now an AI assistant drafts responses automatically, saving an estimated $90K/year.
–75%
Completion time
$90K
Annual savings